The oscilloscope probe is connected to the ground but the signal is AC?

[mike buba]

Hi, please help me to understand what am I seeing on the oscilloscope ... and how to see 0 volts?

I would like to measure a common-mode voltage, i.e. voltage between load neutral and the ground. My setup is shown in the figure below:


However, when I connect only one input of the differential probe to the GND (PE), I see ~ 50 Vpeak, 50 Hz signal on the oscilloscope. The same is if I connect to N (load neutral point) and GND (PE). In both cases, there is no DC voltage at the inverter input.


The multimeter shows zero volts between N and GND.

When I run the inverter, I see this AC voltage superimposed to the common-mode voltage:


I tried with two differential probes: Pico TA057 and Micsig DP10013

[jonpaul]

schematic of setup?

no idea what the connections are probe, scope, mains.

j

[nightfire]

Depending on the mains setup in your area, there is very well the chance that you have a TN network, meaning that usually the N is connected to PE in the main distribution cabinet in your house.
Therefore the voltage between N and PE should (at least in theory) be close to Zero.

[bdunham7]

Where is your oscilloscope grounded?  Is there voltage between your PE and the oscilloscope ground?  (I hope not...)

[Gyro]

According to the OP's diagram, the scope is on an isolation transformer.

[hexreader]

In the UK, ground is typically derived from a metal stake driven into the mud close to the house

Neutral is derived from a ground connection at the electricity sub-station.

It is perfectly normal for up to a few volts AC to be measured between ground and neutral at a wall socket in UK

Right now I have 300mV RMS ground to neutral

At my last house I have measured 8V peak-to-peak (3V RMS)

Each domestic house is supplied with a single phase of a three phase supply, so neutral voltage is likely to vary as the loading imbalance between phases varies

I wouldn't like to guess as to what effects the OPs inverter is causing to the ground and neutral voltages

[bdunham7]

According to the OP's diagram, the scope is on an isolation transformer.

But that doesn't answer my question either way as isolation transformers typically pass the PE through.  If he is floating the scope ground (why!?!) then I'd ask about the power supply for the differential probe as the Y-caps from that may be putting that voltage on the floated components.  And this is a good example of the pitfalls of HV differential probes--they are not isolated! 

[mike buba]

If he is floating the scope ground (why!?!)

I've always been told to flote the oscilloscope ground, especially if I'm

• measuring with multiple grounds and I'm not sure if they are connected
• measuring voltage above 50 V
• combining low voltage (e.g. using probe x100) and PCB measurements

Although now reading about it, it says
Isolation Transformers WARNING! This Is An Unsafe And Dangerous Practice And Should Never Be Done!.

My oscilloscope channels are

• Ch1: x100 votlage (+) to neutral, (-) to earth -> common mode voltage
• Ch2: differetina probe (+) to phase A, (-) to phase B
• Ch3: differetina probe (+) to phase C, (-) to neutral,
• Ch4: current probe

Oscillogram:



I am trying to understand where this AC signal is coming from and how to eliminate it because I am trying to measure common mode voltage, i.e. voltage between load neutral and earth. I have made the connection as shown in this article, Fig. 1, and would expect to measure similar common mode voltage,
Fig. 4. Even this article and this Fig. 12 show Vcm without AC signal superimposed.
When simulating, I get 0 volts between the neutral point and the earth, but in real measurements, there is an AC signal superimposed (see , Ch1)

[themadhippy]

In the UK, ground is typically derived from a metal stake driven into the mud close to the house

you can only guarantee that on a TT system,  TNC-S and TNS  system the spike to earth maybe some distance away

[EPAIII]

Just taking a stab at it.

First, in attachment 2 you are using a differential probe but only one of the two probe leads is attached to anything. The other one is, unassumingly, just floating. It is not unusual for a high impedance, floating input to pick up some radiation from the local power lines. 50 or 60 Hertz RF is floating all around our planet. I have seen this many times when a scope probe is not connected. If you connect that probe to it's common, either directly or with a low value resistor, it should disappear.

Likewise, in attachment 3 I see what looks like the power line frequency and, perhaps, what is the switching frequency of the inverter. Neither of these is surprising at the low, milli-Volt levels of that trace.

In both cases you are seeing only milli-Volts, for gosh sake!

[mike buba]

In both cases you are seeing only milli-Volts

I am using a differential probe x500, so that translates into 100 V or more. Since I am not measuring anything between the natural and the ground with a multimeter, it seems I'm picking something.

Is there a way to eliminate that because it is affecting my other measurements as shown in the oscillograms, where that AC signal is added to a measured common-mode voltage?

[tggzzz]

If he is floating the scope ground (why!?!)

I've always been told to flote the oscilloscope ground, especially if I'm

• measuring with multiple grounds and I'm not sure if they are connected
• measuring voltage above 50 V
• combining low voltage (e.g. using probe x100) and PCB measurements

You have been told a lot of absolute rubbish in multiple ways. Some is dangerous rubbish. Is that perfectly clear?

I suggest you have look at the references in https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/ Those will help you understand

• why your techniques are dangerous - to your scope, to you, and to anybody else in the vicinity
• what class of probe you need to use for different measurements

Hopefully you will then be able to stop using "cargo cult engineering".

EDIT: sorry I was so curt, but the way you have approached dangerous voltages put the wind up me.

There is no substitute for educating oneself from solid reliable sources of information; these are exceptionally easily accessible nowadays, unlike last century. Relying on yootoob vids from random kiddies, or people you chat to over the garden wall is a bad starting point.

The point is not to follow recipes blindly, but to understand why the recipes work. If you can't understand why a recipe works, then it might be that it is merely a "leg of lamb story"[1]. At the very least, compare and contrast multiple recipes for a range of very different sources to get a feel for which "cook" understands and which cook repeats what they do not understand. Deciding which is which is a key skill nowadays.

[1] Child: "Mummy, why do you cut off the end of the leg of lamb before roasting it"?
Mother: "That's the way granny did it when I was young".
Child: "But why?"
Mother: "Go and ask her!"
Child: "Granny, why did you cut off the end of the leg of lamb before roasting it"?
Grandmother, looks puzzled for 30s, then "Oh yes, because we had a small oven and I had to do that to get it to fit in".